ZhenAI

11GB on my macbook & I can't delete it... Does anybody even use it?

@eurofounder things been moving faster recently

The most successful European startup journey: 1. Have an idea 2. Get six co-founders, ideally in their 50s 3. Spend four months to set up a legal entity 4. Apply to Y Combinator, get rejected 5. Write angry LinkedIn post about tech bro culture in the US 6. Raise €50k for 75% equity from top European VCs 8. Do external GDPR audit before users sign up 9. Co-founder leaves to do a second PhD 10. Apply for an EU grant 11. Move headquarters to Estonia for e-residency 12. Launch product, get four users 13. Pivot to sustainability consulting 14. Become the AI innovation advisor to EU Parliament Silicon Valley simply cannot comprehend what we are building here

We launched the "Build your own GPT" from scratch project on NeetCode - inspired by @karpathy 's Lets build GPT series Some reasons you should check it out: - it's beginner friendly / no ML background required - solve 36 real ML problems, learning math, NLP, pytorch, transformers & more - each of your solutions is a .py file - after you solve all problems, you will have built a working GPT clone you can run locally - automatically sync your solutions to github as you progress - a lot of other really cool features, like some problems are inspired by research papers, so for these we embedded the actual paper pdfs and highlighted the important sections, so you dont need to keep opening new tabs TLDR: Learn ML. Build a real project. Show off your work. All 100% free.

@mattpocockuk maaan. this is good. Do you use skills in there or only the .md filed for prompts? Looks elaborate on env management but light on agents/skills/plugins

I built my own software factory, and I open-sourced it. It's called Sandcastle. Here's how to use it:

@ThePrimeagen I wonder when they all gonna switch to SemVer and start releasing model bugfixes too. Oh look at us, we just released mythosV1.2.53, one day after mythosV1.2.52, we are so cool!

I cannot wait for gpt 5.5 here in ~30 minutes

mythos, opus is down, fix immediately, no mistakes, ultrathink ultraplan

WAKE UP

@GergelyOrosz that's the way to go. otherwise I will start a chain of "thinking gyms" in Amsterdam in order for humanity not to lose its ability to... think

There is massive irony in how AI coding tools are starting to become TOO expensive for many enterprises - after eg Anthropic removed subsidizing AI subscriptions. We might go from "everyone use AI for everything!" to "you have $300/month AI budget; use your brain for the rest."

@no_niked phew... drop dead

Soo when Opus 4.5 could one-shot creating an app, Opus 4.7 can one-shot only my limits for Max plan

@ihor_katkov that's bcs you rock! need to do more funky stuff with mr. CC apart of debugging....

@zhenyaThinks Mine is epic!

I am boring...

Claude Code cached a very old file and was very convinced that I am using LLM X in my repo, while I am actually using LLM Y. Spent 2 hours chatting with it living on a wrong assumption...The file was modified months ago though, so how that kind of "mistaken caching" could happen?

@KaiLentit can confirm. mass replaced myself twice while reading this thread

In 2026, AI models expire faster than session cache.

I post this tweet with openclaw. wow. wtf.

@itskerrow yes, go learn something today and apply your skills before it's too late...

@zhenyaThinks the best protection against ai corruption is wanting to run your brain at 100%. intellectual athleticism as defense.

Over-relying on AI can be detrimental to your learning process. Wow. nobody could expect that. What a bummer? 👀 Anthropic conducted a study where junior devs were given tasks to finish with an unfamiliar library. Group A (split into subgroups) with AI and group B - without🧶

I think yes, but we need to keep figuring it out. Check out the study here: anthropic.com/research/AI-as…

On the other hand, delegating more daily puzzles that I am confident enough to solve myself - makes me feel like I'm loosening the grip. Sometimes it takes more time understanding the solution/new feature rather than building it, and the logical question raises: is it worth it?